Eutrophication is the biggest water quality problem threatening society globally (Smith and Schindler 2009, Trends in Ecol Evol 24, 201-207). Increased levels of nutrients released to our rivers and streams due to human activities can cause damaging algal blooms. August algal bloom (Hydrodictyon reticulatum) in Thames basin (Jubilee River)These impair our water supply and harm the environment for wildlife in rivers, estuaries and seas. In the UK, the population is projected to grow in coming decades (by 16% by 2035 in the south-east) and although we can potentially prevent a lot of the water quality problems that this would entail, notably through innovations in wastewater treatment, engineered improvements in urban design and better agricultural practice, we are far less able to control climate-driven effects. We anticipate that in response to global warming eutrophication might become more severe, but we don’t know by how much.

Therefore, a paper recently published in Hydrological Science Journal (HSJ) quantified the effects of future climate in two rivers, the Thames in southern England and the Ure in Yorkshire. To do this, three mathematical models were linked together and daily calculations made over a period of 150 years covering the recent past and projecting through to the end of the 21st Century. Previously, climate models have been shown to produce estimates that are valuable for analysis of future water resources, for example by linking a regional climate model (HADRM3-PPE) with a rainfall-runoff model (CLASSIC), as part of Future Flows Hydrology. To what extent can they also be used to tell us what will happen to the quality of that water? To answer this question, in the new HSJ paper a river water quality model (QUESTOR) was added to these applications.

Model applications showed that by the 2050s there will likely be changes to summer conditions, namely a decrease in river flows due to lower rainfall, and increases in air temperature and solar radiation (less cloudy), with effects likely to be a little more pronounced in the south-east than in Yorkshire. For water quality, setting aside possible effects of any land use change and solely focussing on climate, the number of days when temperature, dissolved oxygen, organic matter (BOD) and algal biomass (chl-a) exceed undesirable values is estimated to increase by between 4.1-26.7 days per year in the River Thames. In contrast in the River Ure, we predict smaller increases in the prevalence of undesirable conditions (by 1.0-11.5 days per year) with some scenarios suggesting no change.

Graph showing expected increases in the number of days per year that water quality thresholds will be exceeded
Above: Expected increases in the number of days per year that water quality thresholds will be exceeded

Models of complex environmental systems are not error-free. Whilst some uncertainty is attributable to rainfall-runoff modelling, water quality simulation was revealed, as expected, to be a greater source, as our understanding of factors controlling river algal blooms is incomplete and a focus for ongoing research. For example, invertebrate grazers such as micro-zooplankton (eg. the rotifer shown in the video below) and invasive zebra mussels can control algal biomass, but it is unclear to what extent. We also were not surprised that the climate model outputs used (namely rainfall, solar radiation and air temperature) proved a source of considerable uncertainty, despite a process of downscaling and, in the case of some variables, bias correction.

Are daily time series of climate model output fit-for-purpose when used for making water quality projections? Probably not, as complications arise when simulating the severity and duration of extreme conditions. Our results show that the climate model produces air temperature and river flow that is unfeasibly extreme in dry summer conditions, critical periods when adverse consequences of eutrophication are most likely. When assessing future water quality and biological impacts, climate model time-series should be used with caution. Alternatively, we recommend use of alternative methods such as change factors (which will be provided as part of the forthcoming UKCP18 dataset) or use of weather generators to produce more plausible daily scenarios.

Video: Brachionus rotifer, collected from the river Thames last year. It is capable of consuming thousands of phytoplankton cells a day, helping to keep algal densities under control.

The paper concludes that future climate alone will bring about a greater prevalence of undesirable river quality by the 2050s, but that accurate predictions are hampered by climatic uncertainty under dry summer conditions when water quality is most vulnerable. This shortcoming needs attention.

This research was in a large part funded by Defra. Previous EA funding helped underpin the use of climate and hydrological models. Two MSc students from Royal Holloway contributed to the testing of the water quality model and the setting up of future scenarios. A NERC-funded PhD studentship (Anna Freeman) under the SCENARIO DTP is investigating biological controls on algal blooms. The work is an example of collaboration in research fostered and coordinated at CEH at a range of levels from policymaking to university postgraduate training. It builds on long-term CEH National Capability research bringing together monitoring and model development.

HUTCHINS, M. G., WILLIAMS, R. J., PRUDHOMME, C, BOWES, M. J., BROWN, H., WAYLETT, A. J., LOEWENTHAL, M., 2016. Projections of future deteriorations in UK river quality are hampered by climatic uncertainty under extreme conditions. Hydrological Sciences Journal. DOI: 10.1080/02626667.2016.1177186

Photograph courtesy of Nicola Ings. Video courtesy of Anna Freeman

Science areas: